In this application for a RCDA (K08), the candidate proposes to obtain expertise in molecular genetics, applying this expertise to the investigation of autism. The candidate is a psychiatrist who recently completed a research fellowship at a Mental Health Clinical Research Center, where he initially performed research in the phenomenology and neurobiology of schizophrenia. Midway through his fellowship, however, driven by his underlying interest in the molecular genetics of neurodevelopmental psychiatric disorders, he began genetic investigations of both schizophrenia and autism. He has now joined the research faculty at his institution and is in a position, with the support of this award, to move towards leading and conducting independent investigations into the molecular genetic basis of these disorders. Autism is characterized by the childhood onset of stereotyped, repetitive behaviors and severe deficits in social interaction and communication. No effective treatment currently exists. Autism is largely heritable, and increasing effort is being expended to identify autism disease genes, though none have yet been discovered. The candidate's Research Plan uses two approaches to attempt to identify such genes: 1) Creating somatic cell hybrids from autistic subjects with chromosome 15 anomalies in order to isolate the abnormal chromosomes, characterize them molecularly, and more narrowly define a potential disease gene harboring region; and 2) Using the genome wide shared segment analysis approach, as well as traditional linkage analysis, to test for linkage in a number of extended pedigrees highly penetrant for autism. Linked or narrowed regions identified through these approaches will be screened for candidate genes, which, once identified, will be evaluated using appropriate molecular techniques. There is also opportunity for the candidate to analyze and compare genetic data from other types of family structures and to examine the effect of novel definitions of the autism phenotype. The candidate's molecular mentoring will be provided by a Howard Hughes geneticist at the University of Iowa, and will be supplemented by mentoring from recognized experts in statistical genetics, genetic computational bioinformatics, the genetics and phenotype of autism, and cytogenetics, and by a rigorous, well- organized series of didactic coursework. Taken together, the training plan and research proposal provide for a breadth of training, the potential for meaningful discovery, and support for the establishment of an independent research career.